1. Limits and Extension of LOD
Position Paper Semantic Summit 2011
Dr. Jörg Wurzer, iQser AG

2. 1.
Limits and Extension of LOD
LOD is a great vision but limited for industry solutions
Limited to standardized RDF data
Today‘s companies distributed data in various formats and silos
Limited to standardized URIs for links
Convention for identifiers are not practicable for distributed corporate data silos
Limited scalability for corporate solutions
Migration of existing data into a triple store multiplies the amount of data
Limited to read access of data
No data production, processing or transactions in real time disqualifies LOD
Shifting the problem of heterogeneity
LOD shifts the problem to the heterogeneity of identifier and ontologies

3. 2.
Limits and Extension of LOD
Extended approach to build smart industry solutions
Extension to any data by efficient semantic integration
Real benefit an potential by mixing up LOD with any existing data with bottom-up-
approach
Extension to dynamic links by semantic analysis
Productively used and enriched LOD by dynamic links of any data beyond RDF and URIs.
Extension to linked actions including graph manipulation
Productively used LOD by linking data with actions to enable new business models.
Extension to scalable persistency concepts
Industry ready by a scalable, lean persistency layer without federation or migration
Extension to uniform information layer
Simplifies access of LOD for existing data silos and formats and adaption of ontologies

4. 3.
Limits and Extension of LOD
Bottom-up versus top-down-approach
Ontologies Data models
Data Model Concept Graph Object Graph
Conventional: Top-down (Ontology) (Linked Concept Graph) (Linked Data Graph)
Manual Mapping
Automatically derived
iQser: Bottom-up
DB DB DB DB
Files Web DB DB

5. 4.
Limits and Extension of LOD
Potential research activities
SPARQL endpoint for uniform information layer
Using a query language standard to query graphs in a complex way to
benefit from the real potential of aggregated and linked data beyond
LOD and RDF.
Performing ontologies on data sets generically
Bridging the gab between abstract domain knowledge and real data by
automatically performed ontologies in an analyzer chain.
Consuming and producing LOD generically
Benefit from mixing up LOD to enrich internal company data and publish
internal data to the LOD cloud for new business models in an
automatically process.

6. Dr. Jörg Wurzer, iQser AG
Member of the Board
joerg.wurzer@iqser.net
www.iqser.com
www.twitter.com/jwurzer
www.youtube.com/iqser